Since a direct precise estimation of parameters with interval data in generalized linear models is generally not possible, Michael Seitz formulates the interval estimates of the parameters as an optimization problem and proposes numerical methods to solve it. The challenge lies in the numerical solution of the high-dimensional optimization problem. This is approximated here using a combination of known numerical methods for non-linear objective functions and heuristic approaches. Additionally, more reliable methods are presented for some special cases.